Details

Autor(en) / Beteiligte

• Han, Zhenhua
• Tan, Haisheng
• Jiang, Shaofeng H.-C.
• Cao, Wanli
• Fu, Xiaoming
• Zhang, Lan
• Lau, Francis C. M.

Titel

SPIN: BSP Job Scheduling With Placement-Sensitive Execution

Ist Teil von

• IEEE/ACM transactions on networking, 2021-10, Vol.29 (5), p.2267-2280

Ort / Verlag

IEEE

Erscheinungsjahr

2021

Quelle

IEEE Xplore

Beschreibungen/Notizen

• The Bulk Synchronous Parallel (BSP) paradigm is gaining tremendous importance recently due to the popularity of computations as distributed machine learning and graph computation. In a typical BSP job, multiple workers concurrently conduct iterative computations, where frequent synchronization is required. Therefore, the workers should be scheduled simultaneously and their placement on different computing devices could significantly affect the performance. Simply retrofitting a traditional scheduling discipline will likely not yield the desired performance due to the unique characteristics of BSP jobs. In this work, we derive SPIN , a novel scheduling designed for BSP jobs with placement-sensitive execution to minimize the makespan of all jobs. We first prove the problem approximation hardness and then present how SPIN solves it with a rounding-based randomized approximation approach. Our analysis indicates SPIN achieves a good performance guarantee efficiently. Moreover, SPIN is robust against misestimation of job execution time by theoretically bounding its negative impact. We implement SPIN on a production-trace driven testbed with 40 GPUs. Our extensive experiments show that SPIN can reduce the job makespan and the average job completion time by up to <inline-formula> <tex-math notation="LaTeX">3\times </tex-math></inline-formula> and <inline-formula> <tex-math notation="LaTeX">4.68\times </tex-math></inline-formula>, respectively. SPIN also demonstrates better robustness to execution time misestimation compared with state-of-the-art heuristic baselines.